308 ALASKAN GLACIER STUDIES
seen from Photo Sta. P., with a photograph (PI. CXXI) by Curtis.1 The amount of advance was accurately determined by Mr. Lewis, our topographer. That the greater part, if not all, of this advance took place after July 1, 1909, is demonstrated by photographs by Grant from Station I (PL CXXII). The central portions of the glacier also advanced somewhat, covering part of the ledges. The southeastern ice cliff was highest in 1910, the fresh-looking cliffs west of the rock ledges next highest, and the northwestern margin lowest of all.
The advance had covered most of the barren zone along the southeastern margin, though some still remained in places. Near the terminus the entire marginal barren zone was covered by ice, and bushes were being overridden, but in front of the glacier there were rock slopes bare of vegetation. Bordering part of the glacier margin was a push moraine of till, bowlders and wood, and there were also some peat rolls. The presence of recently overturned bushes lying on the winter's snow, proves that the advance was in progress during the spring of 1910.
The northwestern margin of Yale Glacier was also advancing rapidly. At the water's edge the glacier terminated on beach and alluvial fan gravels in part undisturbed; but above tide level there was a gravel push moraine including some tree trunks. Farther back along this margin practically all the former barren zone had been overridden, and the glacier was destroying thickets of alder and willow bushes from 10 to 33 years old. The relationship to snow banks proves that part of this advance had taken place during the preceding autumn or early winter; but on July 16, 1910, the advance was still in progress, overwhelming a dense growth of alders and willows, and at a few points rolling up small arches of peat.
Castner Glacier, the southernmost tributary of Yale Glacier, with its pronounced medial moraine and broad lateral moraine, showed no particular change from 1899 to 1910. Higgins' map has this glacier disconnected from the Yale in 1909, but Grant's photographs of that year show this to be an error. This lack of change indicates that the advance of Yale Glacier was not caused by this tributary. The feeders farther back on the east side may have participated in it, for they rise in snowfields which also feed the active northern arm of Meares Glacier in Unakwik Inlet.
Amherst and Adjacent Glaciers. This group of glaciers, on the southeast side of College Fiord has not been studied in detail by any of the expeditions that have visited the region, nor seen from nearer than the eastern edge of the fiord. Amherst Glacier is the largest of the group and both it and Crescent Glacier were shown on a map by Applegate in 1887 and by Gannett in 1899. Williams, Dartmouth and a small glacier to the north, were indicated on the Grant and Higgins map of 1909, as were the Tommy and Cap Glaciers south of Crescent Glacier.
Amherst Glacier is about 4$ miles long and a mile wide and descends from large cirques .near Unakwik Peak. Crescent Glacier is of about the same length but is only about a quarter of a mile wide. Dartmouth and Williams Glaciers are also small. According to Applegate's map, which was made from so near that it is probably correct, Amherst and Crescent Glaciers ended a little nearer the fiord in 1887 than at present and were united.
Gilbert states that in 1899 "Amherst Glacier was passed by the ship at some distance, and its features are known chiefly through the photographs secured by Merriam. It is fed by nave's in full view from the fiord, and approaches the sea in a short, broad stream
i Harriman Alaska Expedition, Vol. HI, 1904, Fig. 43, p, 83.